Abstract
•We examine fitness and host use of hybrid biological control agents from the field.•Phenotypes paired with genomics can assess the risk of non-target attack by hybrids.•Fitness and host use are stable in hybrid and non-hybrid Diorhabda biocontrol agents.•Variation associated with non-target host use may be higher in hybrids.
Releasing several ecotypes or species of biological control agent is common in many biological control programs. However, the consequences of hybridization for fitness and host use of the resulting hybrids are difficult to predict, especially for hybrids between more than two species and with varying levels of introgression. Hybridization of biocontrol agents may increase or decrease fitness, with consequences for control efficacy. Additionally, hybridization may lead to changes in host use patterns that may put non-target host species at increased risk. We collected genomic data to determine ancestry of individuals from ten collection sites across a hybrid zone of Diorhabda spp., the biological control agent of Tamarix spp. in North America. We paired genomic data with phenotypic data on fitness proxies of body size and early fecundity measured in a common garden, and host use for individuals from three of those sites. We found two originally released pure species and a wide range of introgression in hybrids between three Diorhabda species. Body size and early fecundity were similar in pure species and hybrids, indicating hybridization is not detrimental to insect fitness or the biocontrol program and may provide variation that allows populations to become locally adapted. Host use of hybrids was very similar to that of pure species, although some hybrid individuals had increased preference for Frankenia salina, a native non-target species. We find that hybridization has likely not been detrimental to the efficacy and safety of the Diorhabda biocontrol program, but possible impacts on F. salina should be monitored, considering ongoing hybridization and evolution in the field.